It´s an enhanced version of Boxplot.
Boxplots gives a quick view about the structure of a univariate dataset, but remains a black box when it comes to provide clear information about the detailed distribution of this data.
The BeanPlot shows a mirrored estimated density of the distribution, specially usefull to identify multimodal distribution as shown in the picture above (1st serie of data).
This paper provides a good description of Beanplots principles : http://goo.gl/svK8F
And here is the first screenshot of SfE's Beanplot() UDF in progress.... stay tuned.
4 comments:
Looks very interesting. Looking forward to the new version of SparkLines!
Excellent!
Your doing such a nice job man.
This is really good helping people present data in a far better way than was possible before in Microsoft Excel.
No worry about being overruled by new versions of Office, because this add-in will stay at a far higher level.
Really excellent.
This is going to look a little like a small BI pakkage
Just read an article with bean plots and liked the idea ... super, Fabrice! the best, Lars
Post a Comment